Research On Control Method Of Ship Dynamic Positioning Based On Neural Network

With the development of marine resources and water transportation,the demand for the position and attitude of vessels is increasing gradually.The widespread use of the dynamic positioning system has become an inevitable trend.Track following is the main function of the dynamic positioning system,it is required the vessels sail along the designated route smoothly and accurately under the influence of environment such as sea breeze,ocean wave and ocean current.In the dynamic positioning system,the control system is largely effect the track following.This paper taking the marine multi-functional operation ship "299" as the object of study,the starting point is to improve the accuracy of track following,the purpose is researching on dynamic positioning control method based on Neural Network.RBF neural network and CMAC are two typical neural networks which are studied by people so far.They both can be used in the design of dynamic positioning control system.This paper mainly carries on the following several aspects of the work:Established the mathematical model of vessels,studied on the marine environment interference force.Established the mathematical model of sea breeze,ocean waves and currents,their validation is to be conformed through the simulation and test.Researched on RBF neural network and CMAC,put forward RBF neural network and PID compound control method,CMAC and PID compound control method,The simulation results verify the effectiveness of these control methods.The guiding strategy of positioning change is designed with the change of location point coordinate,simulation results verify the feasibility of the proposed method.Use these neural network control methods on the dynamic positioning control system.Designed track controller based on RBF neural network,track controller based on CMAC and CMAC track controller based on RBF identifier.Write these controllers' simulation programs,make the vessels sail along the predetermined route.Compare the control effect of the neural network controller and the traditional PID controller.The rationality and practicability of each route control scheme are analyzed by simulation.Proposed a CMAC control scheme based on RBF identification,the effectiveness of the proposed control scheme is verified by simulation.Designed a CMAC track controller based on RBF identification.The simulation shows that the controller has certain theoretical significance and practical value in the control of dynamic positioning track.This paper make a complete description of the dynamic positioning of the ship's track control process,and description of methods for applying neural network to dynamic positioning ship.It foreshadows the improvement of dynamic positioning system.